Wednesday, August 5, 2009

What price price discovery?

In the financial world, there are some dogmas which participants are keen to reinforce every now and then. The stated justifications for these are mostly on the basis of principle, while the real reasons have generally got to do with more parochial interests (Regulatory arbitrage is one that comes to mind). There are many such dogmas around – the concept of a high-water mark, the notion of insuring bonds to get a higher rating, the idea of low-tax offshore havens, etc. In my mind, the one that has caused the most damage is the notion that liquidity is the holy grail of the capital markets (in fact all markets). We have all been led to believe that liquid capital markets are good for participants, regulators, small investors, governments, everyone else. Why, indeed? Because it aids price discovery. It helps the market know exactly what the market clearing price is for any stock at any point of time.

Creating liquidity in every imaginable market has become a guiding principle for participants that we have often created tradable entities where none should have existed. We want to trade carbon, carbon footprint, grades, virginity?, and many other historical non-tradables. I do not want to enter into the debate of moral questions on trading all kinds of stuff. I am not easily outraged, and more importantly, I would prefer debating on this issue without entering the morality domain. The belief that if anything is tradable, the price would be accurate is ubiquitous. I think this is just plain wrong. Some instruments are meant to be illiquid. In our fetish to create liquidity, we create shallow markets and prostrate before the holy god of price discovery. Merely trading over and over again does not aid price discovery. We merely get to a market-clearing price, and not to a relevant discovered-price.

I think there is an interesting analogy in the science of measurement that might help explain what I am getting at here. There are two interesting terms in measurement – accuracy and precision, both of which give an idea about the efficiency of a measurement but with subtle differences. Accuracy is the measure of how close the measurement is to an actual value, while precision is a measure of how close the measurements are to each other. Accuracy depicts how close we are to the actual quantity. So, in layman terms, accuracy is what is important. Very often, we mistake a precise measurement for an accurate one. According to me, market determined share prices are very precise. Quiet often share prices are “discovered” based on orders accumulated around a point. The more orders that are there, more precise the measurement. But markets do not uncover value accurately. If that were the case, then an index would not move from 8K to 21K in 1.5 years and go all the way back in another 8 months.

Market aficionados will retort saying that there is no “actual” value as far as markets are concerned. Perception is reality, “discovered price” is indeed the actual price. This is a credible argument. One that acknowledges the inherent variability in values of assets based on future assumptions about different factors in the world. But, if this is indeed the case, why do we want precisely discovered prices. Why place the burden of precision on an entity that has no notion of accuracy? The idea of having high levels of precision for an measurement that is inherently inaccurate is absurd. This is like a weather report guy telling you that the temperature for the day should be 23.678 degrees Celsius, but with the caveat that the measurement could be off by 15 degrees on either side. If your range if this broad, your precision counts for pish tosh, a concept with financial markets have never come around to understanding.

If valuing assets is a mugs game, if variance is going to be high, if market dynamics are going to influence prices way more than actual changes, if a change in unemployment rate in the US can take the share price of a local pharma company down sharply, then why should we insist on creating liquidity and aiding price discovery. This piece from Paul Wilmott from New York Times discusses the notion of liquidity and price discovery well.

Leaving aside the question of whether or not liquidity is necessarily a great idea (perhaps not being able to get out of a trade might make people think twice before entering it), or whether there is such a thing as a price that must be discovered (just watch the price of unpopular goods fall in your local supermarket — that’s plenty fast enough for me).

I remember a time when real estate prices of places close to the LSE were higher than ones 200 meters further away – merely for the fact that orders entered into systems in these offices could hit the LSE sooner. We are talking nanoseconds here. If this does not make people look at the financial markets and go – Oh my god, what have we created here? I don’t know what will.

Apparently, this madness is finally being debated. Our fetish for liquidity has led to a scenario where High Frequency Traders account for 50% of NYSE volumes. As ever, this article uses myriad high-falutin words like Flash trading, frequency rebate etc. Finance need not and should not be this complex. If it gets to this level, we should be stopping and restarting.

Nicolas Nassim Taleb wrote a book on randomness in markets called “Fooled by randomness”. The fundamental premise in the entire book is that people do not realise the role randomness plays in life, and in financial markets. That randomness is there is undeniable, and perhaps can be managed. What is more difficult to handle is the fact that inevitably (and periodically) people create structures and thought processes that do not acknowledge the inherent randomness in life (and financial markets).

Why should stock prices trade every day? Why should stock prices be given to the final decimal? We could just have a system that says stock A trades between Rs. 50 and Rs. 55. They announce results on the third Thursday after each quarter ends. The stock can be traded for a week after their results are announced – from the Monday after the results are announced to the Friday. Beyond this, there will be a week in the middle of the quarter when stock A will be open to trade. Prices will still be “discovered”, liquidity will still be there, randomness will be acknowledged, there will be little mark-to-market madness, and insider trading will be non-existent. Mutual funds will not be able to publish NAVs. If there NAVs change from when they publish it to when you read it, the NAV publishing is useless anyway. This is a small price to pay, I would argue. There would be fewer people participating in the price discovery game. This may not be a bad thing.

Weather forecasts often go wrong. The financial market is often like weather forecast. Except that instead of having a weatherman give his thoughts, we have about 25 who contribute their data-points for the daily temperature, and we take the average as a benchmark for the day. Can any of these forecasts change the temperature for the day? Can more estimates improve our accuracy of our prediction? If both these answers are no, then we are probably better off with one guy trying to tie all the dots. All people in the country will have to live with the fact that the temperature could be between 28 and 32 degree Celsius. It is far better than them assuming that it is 30 degree Celsius, when in fact it could be anywhere between 14 and 46.

Imagine the number of weather forecasters freed up to do less damage to the world than they were originally doing. Now, multiply that number by a million. That would probably be the number of resources that can be taken away from harm-creating financial jobs, if the world could shed its fetish for price discovery.

No comments:

Post a Comment